An apparatus for the presentation of visual stimuli at low intensities.

Department of Psychology, Harvard University

William S. Verplanck and Donald S. Blough²

This apparatus provides for flexible control over low-intensity visual stimuli and for accurate recording of the subject’s responses. It permits presentation of stimuli varying in intensity, duration, interstimulus interval, and stimulus patch size, with viewing conditions held constant. Both key pressing and verbal responses may be automatically recorded. A photo-cell timer circuit may be opened by such responses, yielding response latencies. The plans and appearance of the apparatus are shown in Figures 1-5.

A. Optical System (Figure 1)
1. Light SourceThe light source is a strip filament microscope illuminator lamp (G.E. 18A, 1-6V, T-10). A Federal RA-83-A battery charger in parallel with a 6-volt industrial storage battery provides a stable source of power. A Westinghouse PX ammeter monitors the current, which is held at 15 amperes (slightly below the rated amperage of the lamp) by a potentiometer that permits precise control. Cooling air is continuously drawn around their lamp by a fan. After a break-in period of about 10 hours, these lamps maintain constant output over long periods. (The lamp presently in use has been run for about 150 hours without any change measurable with a Macbeth Illuminometer.)

2. Optical PathLight leaves the lamp housing through a one-inch circular stop and passes through a heat-absorbing water cell and condensing system. The image of the filament is brought to focus in the plane of a 11-mm. wide stop located immediately before the blades of the faster shutter. The beam is then projected through (a) the fast and shutters, (b) a collimating lens with stop, (c) a half-silvered mirror which reflects part of the beam to the side and into the aperture of a photocell, (d) the filter and wedge assembly, and (e) the ready and blanking shutters. It is then brought to focus and finally projected on a milk-glass screen located in the wall of the light-tight subject chamber. A portion of this uniformly illuminated milk-glass screen is utilized as the stimulus patch. The light transmission of this system is low; this follows from the space requirements of the various shutters and intensity controls.

The entire optical system is rigidly mounted on a heavy aluminum I-beam. the beam is mounted on wooden four-by-fours, anchored to the concrete floor. A light-tight housing, mechanically independent of the optical system except through the floor, encloses the apparatus and supports all vibration-producing equipment. Two black velvet occluding screens across the housing reduce stray light and reflection. Access to the optical system and controlling apparatus is gained by raising the top of the housing.

Stimulus TimingThe timing of flash duration and interstimulus interval can be put under the control of either of two shutters, the “fast shutter” or the “program shutter.” When one is in use, the other is set in the “open” position, and does not operate.

1. Fast ShutterThe fast shutter is a three-bladed, continuously operating, rotating disc shutter powered by a 1/12 horsepower synchronous Bodine motor. It transmits short, precisely timed flashes very close to square in form. Continuously variable flash durations ranging from 0.0002 to 0.2 seconds may be obtained by suitable adjustment of the apertures of two shutter blades. Onset and offset times of the flash are of the order of .0001 second. These are independent of flash duration. The onset and offset times, and the flash duration were calibrated by the use of a cathode ray oscilloscope and a 1000-cycle oscillator and checked with a Strobotac. The mean flash duration for or all apertures of the faster disk up to 80 degrees was 1.58 times 10 to the negative 4 sec./degree, and the standard deviation was 5. times 10 to the negative 6.

The third blade of this shutter controls the interval between flashes. Interchangable third blades provide interstimulus intervals of 0.58, 1.17, 2.34, 4.67, 9.34, or 18.68 seconds.

Located before each of the apertures in the slow blade of the shutter are pins that close a microswitch at any desired interval from one to three seconds before each flash. The microswitch controls the opening of the “ready” shutter in time for the flash, and it may also activate the “ready” buzzer with the subject’s chamber.

Stimulation by the fast shutter is completely automatic, and control over the presentation of the stimuli to S, either by himself or by E, is activated by the use of keys that control the opening of the “ready shutter.”

2. Program Shutter and ProgrammerThe program shutter provides flashes of relatively long and invariable duration, but gives E complete control of the time elapsing between any two flashes.

The shutter itself consists of a single disk with an aperture of 89 degrees cut from the edge. The disk is driven by a 1 rps Telechron motor. Wiring of the two field coils of the motor is reversed so that when a voltage is impressed across one filed coil, the armature rotates through 270 degrees. When the other coil is similarly activated, a movement of the opposite sign occurs. The direction of movement, then, reverses between successive operations of the shutter. Measures of the flash duration given by this shutter operating with the aperture described above, made photographically, had a mean of 0.270 sec. and a SD of .002 sec. This interval is above the critical duration, so that intensity alone, and not the intensity and duration of flash, controls the visual effect.

A programmer may be used to determine the interstimulus interval for the program shutter. This consists of a punched 35-mm. tape fed under two microswitches as 45 cm. per minute by a synchronous Bodine motor drive. Holes punched along one edge of the tape activate one of the microswitches which, through a stepping relay, causes revolution of the shutter disk. Any series of inter-trial intervals above a minimum of about 0.8 sec. may be punched into the tape. The accuracy of the interval depends on the care used in punching the tape. We have used one “5-second” fixed interval tape which gives an actual mean of 4.93 sec. between flashes (SD of 0.05 sec.).

The other edge of the tape, which passes under the second microswitch, may be punched to control the presentation to the subject of ready signals. The interval between ready signal and stimulus presentation is continuously variable by changing the positions of the two microswitches, one of which slides along a calibrated bar scaled in tenths of a second.

Stimulation by means of the program shutter is alternatively under the direct control of either S or E. In this case, if the subject (or experimenter) presses his “ready” key, the ready shutter opens, and the program shutter passes a single flash.

3. Ready ShutterThe ready shutter permits gross control of stimulus periods, preventing light from reaching the stimulus patch before the start of the experimental series, and during rest periods, regardless of the continuing operation fast or program shutters. It is a metal screen attached to a pivoted, counter-balanced bar. A solenoid lifts the screen from the path of the beam with a click barely audible to S.

The ready shutter opens before each stimulation. It can alternatively be controlled by the fast-shutter ready circuit, the program shutter ready circuit, the experimenter’s ready key, or the subject’s ready key, so that all stimuli can be presented fully automatically , or so that S or E can control which of a series of automatically produced flashes can traverse the whole optical system and so stimulate S. A latching relay holds the shutter open until the subject responds or until the experimenter pushes the “relay reset” button.

4. Blanking ShutterThe blanking shutter is a silent, motor driven disk shutter operated by the experimenter. It may be used to cut off the light without the subject’s knowledge for the presentation of blanks (Vexierversuche).

Intensity and Wavelength ControlsThe color and intensity of the stimulus patch is governed b a set of filters and an optical wedge. As the apparatus has been used, it is equipped with a Baird Associates interference filter (No. 7-2102-A) to limit the light transmitted to a narrow band around 510 millimicrons. Large, step-wise intensity adjustments are achieved by the use of Wratten neutral filters.

Continuous variation of intensity is obtained with an Eastman-Kodak neutral gelatin optical wedge (15 times 2 cm., 2.0 log units, with balancing wedge). The wedge is driven by a fine machine screw which connects, through a gear chain, to a direct-reading counter. Each unit on the counter represents a brightness change of 0.0028 log units.

The wedge may driven either directly by a handle on E‘s control panel, or by a “Zenith” remote-control television tuning motor, which, when activated, clutches to the wedge drive screw. The motor changes the stimulus intensity at a rate of 0.42 log units per second. The motor is controlled by a system of push buttons and limiting switches, which makes possible rapid, automatic setting of the wedge at any one of 12 adjustable stops. The experimenter selects the stop desired and presses the appropriate one of 12 push buttons on his control box. A relay then starts the drive motor, which moves the wedge until a finger on the wedge carriage comes into contact with the stop. A holding relay then closes, and halts the motor until another selection is made. If he wishes, E may move the wedge to any position desired by pressing one of two buttons to determine the direction of the change, and then pressing a master switch to move the wedge as far as necessary. In both cases, a final accurate manual positioning of the wedge drive is negligible with respect to the error term in the photometric calibration of the optical system.

Response Recording
1. Subject’s KeysThree telegraph keys are mounted on a board which fits across the arms of S‘s chair. When a subject presses the left key, a red light for “no” flashes on E‘s panel. A green light, “yes,” flashes when a right key is pressed. The middle key is S‘s shutter key; this key may be inactive, or S may use it to stimulate himself by opening the ready shutter and passing either single or multiple flashes through the program or fast shutters. E may choose among these alternatives by control switches.

Verbal Response SystemAn electronic voice key, with its microphone mounted near the subject’s mouth, utilizes high and low pass filter to separate the high audio frequency predominating in “yes” or “see”³; from the low frequencies of “no.” Thus discriminated, the verbal responses operate “yes” or “no” relays, and are fed into the recording circuit just as if they were produced by depressions of response keys. In the present arrangement, E must reset the voice-key trigger after each response, although there is provision for automat reset by the “ready” circuit.

A Bogen “Challenger” 200M intercommunication system permits E to hear sounds at conversational levels made in S‘s chamber, and hence to hand-record verbal responses of S. S cannot hear E unless E presses a “speak” button.

Stimulus and Response Recordersa. Counters. Two Veeder-Root 110V-DC 8W counter register the total “yes” and total “no” responses, whether they are key depressions or verbal responses through the voice key. A third counter in parallel with the ready shutter, slows the total number of stimulus flashes resented. The counters may be reset at any time by E.

b. Pen recorder. A four-pen polygraph with a tape moving at 1-mm. per second provides a continuous record of stimuli, responses, and the time relations in which they occur. Three of the pens are wired in parallel with the counters described above so that they record stimulus presentations, “yeses,” and “noes.” The fourth pen is activated when the blanking shutter is closed; this record permits enumeration of blanks and false responses.

c. Response-latency timer. The time elapsing between the presentation of a flash and a response may be read from a Standard Electric Time Co. S-1 electric stop-clock calibrated in hundredths of a second. Part of the light passes by either the fast shutter or the program shutter is reflected away at 90 degrees from the axis of the optical system by a half-silvered mirror mounted in the path of a beam. This reflected light enters a photocell which trips a trigger circuit that closes a latching relay. This relay engages the clock clutch. The clock stops when current from the response circuit unlatches the relay. The photocell circuit responds consistently to flashes as little as 0.002 sec. in duration. The circuit lag between the time of the flash and the closing of the clock clutch is 0.0013 sec. with SD of 0.0021 sec. (one hundred readings). These constant and variable errors are small with respect to the durations measured.

Viewing Conditions; S‘s ChamberThe subject chamber is a 200 by 125 by 156 cm. light-tight box. It is accessible through a large double door which is held firmly closed with springs and latches, but which can be readily opened by the subject from inside. A fan draws a continuous steam of air through the chamber from a duct near the floor to a vent in the ceiling.

1. Headrest (Figure 5)The headrest holds the head of S in a fixed and reproducible position with the use of a biting board (not shown in plate). Three points of contact, one at the chin and two on the forehead, hold the head steady and comfortable. The headrest is secured in a metal frame anchored rigidly to the ceiling and walls of the S chamber. It has several adjustable dimensions:

a. Pupil diameter. Pupils with large apertures (6.40 mm. in diameter) are shown in figure 5. These may be readily replaced with pupils of other diameters.

b. Interpupillary distance. The knob shown adjusts for any IPD from 55.0 mm. to 74.0 mm. ± 0.2 mm.

c. Angle of plane of pupils to the vertical. This is shown set at 10 degrees from the vertical, so that the visual axes are 10 degrees below the horizontal plane. It is adjustable from 0 degrees to 15 degrees, ± 20′. This angle remains fixed for all S‘s.

d. Horizontal base distance: adjustable.

e. Angle of brow rest from vertical: adjustable.

f. Height of mouth-bite: adustable.

g. Protrusion of mouth-bite: adjustable.

h. Height of chin rest: adjustable.

i. Microphone position: adjustable.

The position of a point midway between the artificial pupils is fixed with respect to the stimulus patch, and all adjustments for individual S‘s are made with respect to this point. The values on items 2, 4, 5, 6, 7, and 8 are set with each Sfor maximum clarity of the stimulus patch and fixation point at high brightness. Once the values for a given S are determined, they are recorded, and they may quickly be reproduced on the headrest before each run with that S.

2. Fixation Point and Stimulus PatchThe fixation point is a 1-mm. hole in a metal plate fixed across the end of a small tube. Light from a neon glow lamp enclosed in the tube is attenuated by several layers of Polaroid X-DA-12 red dark-adaptation plastic, which are placed between the lamp and the hole. The fixation light may burn continuously, or it may be activated by the ready circuit and shut off by S‘s response.

The size of the milk-glass stimulus patch is determined by a circular hole in a thin metal diaphragm that slides in front of the milk-glass projection screen. Plates with various apertures may be used. We have been using an aperture 2.79 cm. in diameter that subtends a visual angle 3 degrees 0’ at the cornea.

Other Apparatus in S’s ChamberThe ready signal is sounded by an ordinary door buzzer located in an upper corner of the chamber. the auditory characteristics of the signal may be changed by adjusting the armature of the buzzer.

The intercom loudspeaker is located about 30 cm. below the stimulus patch.

The voice-key microphone and S‘s chair, with its ready and response keys, are the remaining items in the chamber.

The walls and all equipment in the chamber are painted flat black, with the exception of certain parts of the headrest.

Auditory Conditions in the Subject ChamberThe fan that ventilates the subject’s chamber produces a continuous sound level of 65 db. in the chamber (measured with a G.R.C. T-759 sound level meter); it partly masks extraneous sounds. Above this S can hear the click from the program shutter before each flash. The opening of the ready shutter can also be heard if the ready buzzer is switched off. Another auditory cue is produced by a wedge drive; the duration of this sound is correlated (but not in a simple manner) with the change in light intensity produced by the movement of the wedge. A second fan-motor may also be placed in the chamber. When this is used as a further source of masking noise, apparatus sounds cannot be distinguished.

Under ordinary conditions, S reports that he hears no noise from the experimental room outside the chamber during experimental runs. However, loud sounds can be heard if they occur. For example, conversation is audible, but it is understood only with difficulty.

F. Experimenter’s Desk and Control PanelsDuring an experimental run, E sits at the table shown in Figure 3, with response recorders, wedge indicator, and the most frequently used controls close at hand. The top of the apparatus housing need not be opened during an experimental session. Before the session E may raise the top in order to add or remove filters and to make sure that unused shutters are in the “open” position.

Indicator lights sow whether various units of the apparatus are on or off (e.g., lamp, shutter drives, wedge drive). Lights also indicate whether the ready and blanking shutters are open and in which direction the wedge is set to move when next activated.

Footnotes* Received in the Editorial Office on August 7, 1953.
1 Construction of the apparatus described in this paper was begun in the Department of Psychology, Indiana University, under ONR Contract N6onr-180, T.O. IV, Project 143-253, and with support from a grant by the Graduate School of arts and Sciences. Work is continuing at Harvard University under Contract N5ori-07639.
2 We wish to acknowledge with gratitude the contributions of Dr. George H. Collier, Duke University; Mr. John R. Binford, Indiana University; Mr. John W. McCrary, Brown University; and Mr. James A. Seddon, Jr., Harvard University, to the design and construction of various features of the apparatus; and of Dr. Lorin A. Riggs, who designed and constructed the fast shutter.
3 The subject may be instructed to respond “see,” rather than “yes,” because the s sound at the beginning of the word is preferable if reaction times are measured.